epa’s benefit cost analysis - home - energy and...
TRANSCRIPT
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 1
EPA’s Benefit Cost Analysis
Air Quality Division
Susana Hildebrand, P.E., Chief Engineer
Michael Honeycutt, Ph.D.
Stephanie Shirley, Ph.D.
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 2
Texas Commission on Environmental Quality
Mission Statement: The Texas Commission on Environmental Quality strives to protect our
state's human and natural resources consistent with sustainable economic development. Our goal is clean air, clean water, and the safe
management of waste.
The TCEQ regularly weighs matters that affect the environment and economy. Our goal is sensible regulation that addresses real environmental risks, while being based on sound science and
compliance with state and federal statutes. In every case where Texas disagrees with EPA’s action, it is because EPA’s action is not consistent
with these principles.
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 3
Background
• March 2011 – EPA published “Benefits and Costs of the Clean Air Act from 1990 to 2020 (Second Prospective Study)”
– Benefits ($2T) outweigh costs ($65B) by 30 to 1
TCEQ staff examined this analysis, focusing on:
• The studies used
• The assumptions made
• The methods employed
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 4
Regulatory Impact Analyses
• President requires RIAs (Regulatory Impact Analyses) from all agencies proposing significant regulations
• RIA should help determine if the benefits of an action are likely and justify the costs or discover which of various possible alternatives would be the most cost-effective
– (OMB circular A4, 09/2003)
• RIAs are NOT subject to peer or public review
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 5
Key legislation – Executive Orders
• EO12291 – Reagan, 1981 – “Regulatory action shall not be undertaken unless the
potential benefits to society for the regulation outweigh the potential costs to society…the alternative involving the least net cost to society shall be chosen”
• EO12866 – Clinton, 1993
– Key change: benefits must justify the costs
• EO13563 – Obama, 2011
– Benefits must justify the costs
– New: equity, human dignity, fairness and distributive impacts are required to be considered
– “Our regulatory system must protect public health, welfare, safety, and our environment while promoting economic growth, innovation, competitiveness, and job creation”
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 6
Use of PM2.5 in RIAs
From Smith, 2012 testimony
2009 Change in
Methodology
• EPA uses estimates of benefits from reducing PM2.5 in its RIAs for rulemakings under the Clean Air Act • This is called “co-benefits”
because a PM2.5 reduction is expected from efforts to reduce other air pollutants
• Trend towards using PM2.5 as primary source of benefits in most RIAs since 1997 • Even when regulation is
not intended to protect public health from exposures to ambient PM2.5
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 7
Key Changes in PM2.5 Methodology
• The Benefits and Costs of the Clean Air Act from 1990 to 2020 (March 2011)
1. A no-threshold model for PM2.5 that calculates incremental benefits down to the lowest modeled air quality levels
2. Risks attributed to very low (background) levels of ambient PM2.5
3. Assumption of causal relationship between PM2.5 and
mortality
4. A Value of Statistical Life (VSL)
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 8
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
Pre-2009 Post-2009
Number of Deaths due to PM2.5 in 2005
Result of Key Changes in PM2.5 Methodology
Change in deaths attributable to PM2.5
Increased estimates of benefits
88,000 4% of all deaths
in U.S.
Despite improvement in air quality since the CAAA
Source: EPA 2010 Quantitative Health Risk Assessment for
PM2.5 Table G-1
320,000 13% of all deaths
in U.S.
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 9
1. No Threshold Model
• A no-threshold model for PM2.5 that calculates
incremental benefits down to the lowest modeled air quality
levels
Adapted from
Smith 2011
Annual average for
city – estimated
from monitors
Based on
death
certificates
collected in
the city
Annual
NAAQS
level
Average Annual PM2.5 Eftim et al. 2008
Mo
rta
lity R
isk
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 10
1. No Threshold Model
• A no-threshold model for PM2.5 that calculates
incremental benefits down to the lowest modeled air quality
levels
Adapted from
Smith 2011
Statistically fitted
concentration-response
function
Annual
NAAQS
level
Average Annual PM2.5 Eftim et al. 2008
Mo
rta
lity R
isk
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 11
1. No Threshold Model
• A no-threshold model for PM2.5 that calculates
incremental benefits down to the lowest modeled air quality
levels
Adapted from
Smith 2011
Statistically fitted
concentration-response
function
Extrapolation
below lowest measured levels
Annual
NAAQS
level
Average Annual PM2.5 Eftim et al. 2008
Mo
rta
lity R
isk
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 12
1. Question: what is the shape
of the curve in the low-dose
range?
1. No Threshold Model
• A no-threshold model for PM2.5 that calculates
incremental benefits down to the lowest modeled air quality
levels
Adapted from
Smith 2011
Statistically fitted
concentration-response
function
Extrapolation
below lowest measured levels
2. Question: is there significant
risk associated with ambient
PM2.5 levels?
threshold
Annual
NAAQS
level
Average Annual PM2.5 Eftim et al. 2008
Mo
rta
lity R
isk
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 13
January 2010 – June 2011
41 Volunteers
Dose:35 – 750 ug/m3
Results:
1 individual: elevated heart rate
1 individual: irregular heart beat*
39 individuals: no clinical effects
Clinical Exposure Studies Conducted by EPA
* Case Report: Supraventricular Arrhythmia after
Exposure to Concentrated Ambient Air Pollution
Particles. Ghio et al. EHP. Feb. 2012. 120:275-277
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 14
2. Risk Attributed to Ambient PM2.5
NAAQS 15µg/m3
≈99% of the estimated mortality is due to concentrations less
than the level deemed protective of public health (NAAQS). Deaths due to “unsafe”
PM2.5 levels
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 15
3. Assumption of Causality
No Effect
Relative Risk
for Smoking
and Death from
Cardiovascular
Disease
Relative
Risk for
Smoking
and Death
from Lung
Cancer
Relative Risk
Scientific & Legal
Guidance
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 16
Dallas
Birmingham
Las Vegas
Houston
Riverside
Cincinnati
Los Angeles
Seattle
Washington DC
Palm Beach
Detroit
Pittsburgh
San Diego
Fresno
Philadelphia
Sacramento
Indianapolis
Manhattan
Minneapolis
Cleveland
Boston
Tampa
Columbus
Memphis
Chicago
Phoenix
Milwaukee
Weibull
Distribution (which cannot show
zero or negative
relationships)
3. Assumption of Causality
PM 2.5 -Mortality Coefficient Estimates and 95% CI
Adapted from Franklin et al. 2007
Estimates of the percent Increase in all-cause mortality with a
10 µg/m3 increase in previous day’s concentration PM 2.5
• The epidemiology studies cannot show causality
• The analysis “assumes a causal relationship between PM2.5 exposure and premature mortality…if the PM2.5/mortality relationship is not causal, it would lead to a significant overestimation of net benefits”
-EPA, The Benefits and Costs of the Clean Air
Act from 1990 to 2020, March 2011
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 17
Extrapolation of Mortality Estimates
From EPA – Regulatory Impact Analysis of the Proposed
Toxics Rule: Final Report – March 2011
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 18
4. Value of Statistical Life Definition
• A Value of Statistical Life (VSL) = value of risk reduction
A “statistical life” has traditionally referred to the aggregation of small risk reductions across many individuals until that aggregate reflects a total of one statistical life
The VSL has been a shorthand way of referring to the monetary value or tradeoff between income and mortality risk reduction, i.e. the willingness to pay for small risk reductions across large numbers of people
It has led to confusion because it has been interpreted as referring to the loss of identified lives
If risk was reduced by 1 in 1,000,000
for 1 year
in a population of 200 million savings of 200 actual lives
savings of 200 statistical lives = value of risk reduction
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 19
Deriving Value of Statistical Life Willingness to Pay – Road Hazard Studies
• Example:
– Cars with seatbelts cost $300
more than cars without seatbelts
– Buying a car with that option
reduces the probability of death
by 1 in 100,000
– If people are willing to pay for this option, we can infer that the person is placing a valuation on his/her life of at least $300 x 100,000 = 30,000,000 ($30 million)
$300
Probability of death by 1 in 100,000
$300
x 100,000
= $30 million
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 20
Deriving Value of Statistical Life Income vs. Risk – Occupational Studies
• Example:
– A job carries a higher risk of
injury, but pays $ 500 more per year
– The more dangerous job carries
an increased risk of injury by
1 in 10,000
– If people are willing to pay for this option, we can infer that the individuals are placing a valuation on
their lives of at least $500 x 10,000 = 5,000,000 ($5 million)
$ 500
x 10,000
= $5 million
$ 500
Probability of injury by 1 in 10,000
vs
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 21
Interpreting VSL in the Media
“When these new [EGU MACT] standards are finalized, they will assist in preventing 11,000 heart
attacks, 17,000 premature deaths, 120,000 cases of childhood asthma symptoms and
approximately 11,000 fewer cases of acute bronchitis among children each year. Hospital visits will be reduced and nearly 850,000 fewer days of work will be missed due to illness.”
- Lisa Jackson, EPA Administrator, 2011
This was interpreted as:
“EPA’s proposed mercury and air toxics standards … are projected to save as many as 17,000 American lives …
- John D. Walke, Natural Resources Defense Council, 2011
“These new standards mark a huge step forward in clean air protections and will be responsible for saving thousands of lives each year.”
- Albert A. Rizzo, MD, National Volunteer Chair of the American Lung Association
“The new EPA mercury standards will save countless lives and improve the quality of life for
millions.”
- New York Mayor Michael Bloomberg
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 22
Appropriate Use of Value of Statistical Life
• Lives Saved vs. Life-Years Added
– Deaths “prevented or avoided”
– Gains in life expectancy
• The median age of people who gain extra months of life from cleaner air is close to 80 years
• Adjustment of VSL for quality of life:
– EPA VSL of $8,900,000 appropriate for healthy young adult (≈25)
– 6:1 ratio for 25 vs. 80 year old
From Weeks 1995
EPA VSL:$8,900,000
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 23
Clean Air Act - Benefits and Costs
life-years gained in 2020 * value per statistical life-year gained
=1,900,000 life-years gained * $150,000/life-year gained
≈ $0.3 trillion
Benefit/Cost = $0.3 trillion/$0.065 trillion* ≈ 5
reduced number of deaths in 2020 * value per statistical life saved
= 230,000 fewer deaths * $8,900,000 per life saved
≈ $2 trillion
Benefit/Cost = $2 trillion/$0.065 trillion* ≈ 30
Adjusted estimate of benefit:
$19 billion
Benefit/Cost = $0.019 trillion/$0.065 trillion* ≈ 0.3
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 24
Mercury & Air Toxics Standard
Benefits from HAPs (billions)
“Co-Benefits” from non-HAPs (billions)
Mercury $ 0.004-0.006 $ 1-2
Acid Gasses $ 0 $ 32-87
Non-Hg Metals $ 0 $ 1-2
Total ≤$ 0.006 $ 33-90
Table adapted from testimony by Anne E. Smith 2/2010 to Subcommittee on Energy and Power
• MATS is estimated to prevent 0.00209 IQ point loss per child (starting immediately)
• Each child will gain 0.0956 school days over their lifetime
• 0.00209 IQ points x 244,468 children = 511 IQ points per year
• Assuming a net monetary loss per decrease in one IQ point of between ~$8,000 and ~$12,000 (in terms of foregone future earnings)
• Benefit = $4.2M to $6.2M
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 25
Oil & Gas NSPS and NESHAPS
“…quantification of those benefits cannot be accomplished for this rule. This is not to imply that
there are no benefits of the rules; rather, it is a reflection of the difficulties in modeling the direct and
indirect impacts of the reductions in emissions for this industrial sector with the data currently available.” April 2012 RIA
Oil and Natural Gas NSPS (millions)
Oil and Natural Gas NESHAP Amendments
(millions)
Benefits NA NA
Costs - $15 $3.5
Non-monetized benefits
11,000 tons of HAP5 190,000 tons of VOC
1.0 million tons of methane Health effects of HAP exposure
Health effects of PM2.5 and ozone exposure Visibility impairment Vegetation effects
Climate effects
670 tons of HAP 1,200 tons of VOC
420 tons of methane Health effects of HAP exposure
Health effects of PM2.5 and ozone exposure Visibility impairment Vegetation effects
Climate effects
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 26
PM Co-Benefits in RIAs
• Double counting benefits: same statistical lives counted in multiple rules
• Different costs: unique to each rule
PM2.5 NAAQS
Utility Boiler MACT
Mercury Air Toxics Standard
Sewage Sludge
Incineration
Units
Ferroalloy NESHAP Total
Costs millions
($2006) Estimated Statistical
Deaths 15,000 11,900 2,650 25 14
Cost 6,400 10,600 9,329 17 4 26,350
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 28
Contact Information
Susana M. Hildebrand, P.E.
Chief Engineer
(512) 239-4696
Michael Honeycutt, Ph.D.
Division Director, Toxicology
(512) 239-1793
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 29
Health Effects of Poverty and Unemployment
• Poverty and unemployment have been recognized as risk factors for morbidity and mortality since the 1800’s (Virchow, 1848)
• As of March 2012, there are 4,850 publications on this topic
Roelfs et al. Soc Sci Med 2011; 72:840-54
Relation of real GDP per capita to age-adjusted death
rates, US 1900–2000 (natural logarithms).
Brenner M H Int. J. Epidemiol. 2005;34:1214-1221
Unemployment and All-Cause
Mortality
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 30
With CAAA vs. Without CAAA
•Freezes pollution controls at 1990
levels
•Assumes no additional state or local
regulation after 1990
•Assumes no improvements in
technology or efficiency
•“There is no way to validate the
counterfactual, without-CAAA
scenario estimates”
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 31
Oil & Gas NESHAPS
April 18, 2012 Press Conference
“Today’s rules would yield significant reductions in methane, a potent greenhouse gas. EPA’s
Regulatory Impact Analysis for the rule estimates the value of the climate co-benefits that
would result from this reduction at $440 million annually by 2015.”
-Gina McCarthy
Note: benefits calculated at 3%, but costs at 7%
Reported monetized benefit: $0
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 32
Costs of the Clean Air Act and Amendments
• Cross State Air Pollution Rule
– EPA estimated cost:$800 million annually
– Independent analysis: $120 billion by 2015
• Boiler MACT
– EPA estimated cost:$2.6 billion annually
– Independent analysis:$14.5 billion
Year RIAs for Rules Not Targeting Ambient PM 2.5 PM Co-
Benefits are >50% of Total
PM Co-Benefits Are Only Benefits
Quantified
Cost ($ Billion)*
1997 Ozone NAAQS (.12 1hr=>.08 8hr) x 9.60
1997 Pulp&Paper NESHAP 6.48
1998 NOx SIP Call & Section 126 Petitions 1.66
1999 Regional Haze Rule x 1.74
1999 Final Section 126 Petition Rule x 1.15
2004 Stationary Reciprocating Internal Combustion Engin NESHAP x 0.25
2004 Industrial Boilers & Process Heaters NESHAP x x 0.86
2005 Clean Air Mercury Rule x 0.90
2005 Clean Air Visibility Rule/BART Guidelines x 1.50
2006 Stationary Compression Ignition Internal Combustion Engine NSPS 0.06
2007 Control of HAP from mobile sources x x 0.36
2008 Ozone NAAQS (.08 8hr =>.075 8hr) x 8.20#
2008 Lead (Pb) NAAQS x 3.20
2009 New Marine Compress'n-Ign Engines >30 L per Cylinder x 1.90
2010 Reciprocating Internal Combustion Engines NESHAP - Comp. Ignit. x x 0.37
2010 EPA/NHTSA Joint Light-Duty GHG & CAFES 15.60
2010 SO2 NAAQS (1-hr, 75 ppb) x >99.9% 1.50
2010 Existing Stationary Compression Ignition Engines NESHAP x x 0.25
2011 Industrial, Comm, and Institutional Boilers NESHAP x x 0.49
2011 Indus'l, Comm'l, and Institutional Boilers & Process Heaters NESHAP x x 2.90
2011 Comm'l & Indus'l Solid Waste Incin. Units NSPS & Emission G'lines x x 0.28
2011 Control of GHG from Medium & Heavy-Duty Vehicles 2.00&
2011 Ozone Reconsideration NAAQS x 8.20#
2011 Utility Boiler MACT NESHAP (Final Rule’s RIA) x ≥99% 9.60
2011 Mercury Cell Chlor Alkali Plant Mercury Emissions NESHAP x 0.00
2011 Sewage Sludge Incineration Units NSPS & Emission Guidelines x x 0.02
2011 Ferroalloys Production NESHAP Ammendments x x 0.004
Total: 60.67 + MATS – 9.3 Partial Total: 69.97
* ($2006)
CAAA Benefit Cost Analysis • TCEQ Chief Engineer’s Office • May 17, 2012 • Page 34
Adjusted Benefits Estimate
Tony Cox, 2011:
($1.8 trillion initial estimate)
x (1/6 reduction factor for VSL if age or VSLY is considered)
x (0.5 probability that a true association exists)
x (0.5 probability that a true association is causal, given that one exists)
x (0.5 probability that ambient concentrations are above any thresholds or nadirs in the C-R function, given that a true causal C-R relation exists)
x (0.5 expected reduction factor in C-R coefficient by 2020 due to improved medication and prevention of disease-related mortalities)
= (1.8 trillion)*(1/6)*(0.5)*(0.5)*(0.5)*(0.5) = $19 billion